Two new hypotheses suggest that the key pathology in migraine has a neuronal origin. A pivotal role is assigned to brain hypoxia (1) and spreading depression (SD) (neuronal depolarization spreading gradually over the cortex) (2). Flunarizine has been tested both against brain hypoxia and SD. Its potent antihypoxic properties in animal models led to its use as a prophylactic drug in migraine therapy. Earlier experiments suggested that flunarizine shortened recovery after neuronal depolarization. Recent experiments suggest that flunarizine may enhance the threshold for the elicitation of SD. Finally, it is often unclear whether the effects observed with flunarizine are due to a vascular or a direct neuronal effect. Therefore, a study was made to show whether flunarizine affected hypoxia-induced alterations in synaptic function in slices of hippocampus held in vitro. At physiological drug concentrations in brain, flunarizine improved post-hypoxic recovery of synaptic function. A direct neuronal protective effect was thus demonstrated.